There are an estimated 200,000 HIV-serodiscordant heterosexual couples in the U.S. and the majority of heterosexually acquired HIV infections each year (>12,000) can be attributed to transmission within these couples. Only about one-third of serodiscordant couples use condoms consistently and less than half of infected partners maintain undetectable viral loads. HIV pre-exposure prophylaxis (PrEP) represents an efficacious and promising new approach to reduce the spread of sexually acquired HIV among serodiscordant couples. In 2012, the FDA approved daily oral FTC/TDF (Truvada) as the first PrEP agent for HIV prevention indication. The CDC has identified HIV-uninfected persons in serodiscordant relationships as a priority for PrEP in the U.S. Yet, this group has not been included in prior PrEP clinical trials or demonstration projects and many critical questions remain unanswered regarding the levels and determinants of PrEP acceptability, medication adherence, health effects, clinical monitoring, changes in sexual behavior, and retention in care. This fundamental gap in knowledge represents a significant barrier to realizing the public health benefits of PrEP. To fill this gap, we propose a mixed-methods prospective open-label demonstration project of daily oral Truvada/PrEP among HIV-serodiscordant heterosexual couples in high-prevalence communities in New York City (NYC). Roll-out of oral PrEP in NYC (and nationally) is at an early stage and health providers and other stakeholders are just beginning to confront and propose solutions to implementation challenges. This provides an historic opportunity to study the implementation process across clinical sites and identify factors at multiple levels (provider, couple, individual) associated with PrEP outcomes. HIV-negative PrEP initiators (n=270) and their HIV- positive index partners (n=270) will be recruited through a network of clinical sites (N=18) and followed for 18 months with assessments at 3-month intervals. PrEP implementation data will be collected from clinical providers and behavioral, biological, and medical record data will be collected from PrEP users and index partners. Data will also be collected from a sample of those who do not initiate PrEP (n=100). Qualitative interviews will be conducted with a subset of providers and couples. The specific aims are to describe the temporal patterns and identify multilevel determinants of PrEP outcomes, including (a) uptake, acceptability and retention in care, (b) medication adherence, (c) safety and toxicity, (d) compliance with HIV testing and clinical monitoring, and (e) changes is sexual risk behavior. This research is innovative in that it employs a biobehavioral multi-site longitudinal design to identify critical factors affecting PrEP implementation outcomes at multiple levels-provider (e.g., policies, procedures), dyadic (e.g., partner and relationship attributes), and individual (eg., PrEP user characteristics)-in an understudied group. The significance and potential impact of this study is in its capacity to advance the field by informing policy, advancing evidence-based clinical practice, guiding the development of interventions, providing data for PrEP modeling, and directing future research on models of PrEP delivery.